https://bz.apache.org/bugzilla/show_bug.cgi?id=61733

--- Comment #5 from Rainer Jung <rainer.j...@kippdata.de> ---
Created attachment 35549
  --> https://bz.apache.org/bugzilla/attachment.cgi?id=35549&action=edit
Reorder update actions and introduce flag for demanding updates for all members

I think both of your analysis are right:

- we only sync via shared memory lb_mult for the worker whose factor changed
- the order of steps in case only one member is changes is wrong

The attached patch should fix both issues, correct the order - similar to what
Mark suggests - plus letting jk_lb_push know, that it should not only push none
or one member, but instead needs to push all members of an lb. To make this
possible we change the signature of jk_lb_push and introduce a new flag.

The patch compiles, but I'm short of time to aczually test it. it would be
great if one or both of you could give it a try.

Comments very welcome.

-- 
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org

Reply via email to